Exit Polls

[Tender Branson]

So, after the news media said we were getting only 31 state exit polls and then changed it to 50+1 state exit polls, we now ended up with the 31 states ?

Or did I miss the other state exit polls ?

(only 31 states)

http://www.foxnews.com/politics/elections/2012-exit-poll

All other news media (CNN, CBS etc.) also have only 31 states released.

[greenforest32]

I was running into this too. I can't find exit polls for all 50 states...

You must be logged in to read this quote.

How can they live with themselves?

[Sbane]

I think the North Dakota Senate race showed why exit polls in all states are necessary.

[Ljube]

I'm afraid, this is the main reason Romney lost:


Who is more to blame for current economic problems?

	Total	Obama	Romney
Barack Obama	38%	5%	94%
George W. Bush	53%	85%	12%

[True Federalist (진정한 연방 주의자)]

They only relented enough so as to do enough exit polling in the other 19 states + DC so as to be able to do a representative national exit poll.  However, the sample sizes on those other 20 places were to be too small to have a decent MoE, so they were not planning on releasing state-level data for them.

[Tender Branson]

They had a national sample of 25.000 !!!

Texas ? Too small ?

It would yield a sample of 2100 for Texas.

I can understand Wyoming, which would only have a sample of 40 people. But Texas ? Georgia ?

[True Federalist (진정한 연방 주의자)]

Who said they polled each state in proportion to its population?

[Tender Branson]

Who said they polled each state in proportion to its population?

What sense would it make to poll 31 states with big samples and only poll a handful of people in the other 19 states ? If they use 1 set of standard for the 31 states and another set for the other 19, the national exit poll could be flawed ...

[True Federalist (진정한 연방 주의자)]

Who said they polled each state in proportion to its population?

What sense would it make to poll 31 states with big samples and only poll a handful of people in the other 19 states ? If they use 1 set of standard for the 31 states and another set for the other 19, the national exit poll could be flawed ...

And of course, we know that pollsters never use flawed models.  But more seriously, in the case of Texas, I doubt they sampled more than a couple of representative precincts and then scaled the results based on the actual vote.  Not great, but better than no exit poll at all.

[Hash]

The lack of exit polls in big states like TX and GA (and a bunch of other states) is the biggest tragedy of 2012.